Share Email Print

Proceedings Paper

Maximum constrained sparse coding for image representation
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Sparse coding exhibits good performance in many computer vision applications by finding bases which capture highlevel semantics of the data and learning sparse coefficients in terms of the bases. However, due to the fact that bases are non-orthogonal, sparse coding can hardly preserve the samples’ similarity, which is important for discrimination. In this paper, a new image representing method called maximum constrained sparse coding (MCSC) is proposed. Sparse representation with more active coefficients means more similarity information, and the infinite norm is added to the solution for this purpose. We solve the optimizer by constraining the codes’ maximum and releasing the residual to other dictionary atoms. Experimental results on image clustering show that our method can preserve the similarity of adjacent samples and maintain the sparsity of code simultaneously.

Paper Details

Date Published: 14 December 2015
PDF: 7 pages
Proc. SPIE 9813, MIPPR 2015: Pattern Recognition and Computer Vision, 98130V (14 December 2015); doi: 10.1117/12.2204911
Show Author Affiliations
Jie Zhang, Beihang Univ. (China)
Beijing Key Lab. of Digital Media (China)
Danpei Zhao, Beihang Univ. (China)
Beijing Key Lab. of Digital Media (China)
Zhiguo Jiang, Beihang Univ. (China)
Beijing Key Lab. of Digital Media (China)

Published in SPIE Proceedings Vol. 9813:
MIPPR 2015: Pattern Recognition and Computer Vision
Tianxu Zhang; Jianguo Liu, Editor(s)

© SPIE. Terms of Use
Back to Top